… and hate living it now.
(Incidentally I don’t recall the bit with Stuart Russel at the end being part of this when I first watched it, and feel it dilutes its power now: there is no reason to say “Given developments in A.I. and drones, someday soon this is going to be real!” It’s already real; it’s called guns: 1 out of every 20 Americans owns an AR-15; 3 in 10 own a gun of some sort. Only half of those guns are stored under lock and key, and only a third unloaded.)
… ’cause that’s pretty fucked up, man.
Rough chuckles from our friends at The Onion, but they ain’t wrong:
Remember this project from last November?
Well, for those who like closure:
Y’all did it!
Just before Xmas we closed out the project, having raised a bit more than $10,000 with your help. The end result was the abolition of $1,114,133.50 in shitty medical debt, freeing 2,006 Michiganders across 17 counties from one more dumb late-stage capitalism headache:
Bummed that you missed out on this project? Then I have excellent news:
You and your pals can start your own! The folks at RIP Medical Debt are super sweet and helpful, and it’s a blast.
I want that to be the takeaway here: there are plenty of religious and moral and political and ethical reasons to work on a project like this, but I still believe the best reason to do it is that it is fun. Every day, multiple times each day, it would occur to me: We’re going to nuke a million dollars of medical debt. And just that thought, that a bunch of us regular folk going about our regular lives could do just a little something more and move such a big goddamn rock—it’d make me laugh. It literally tickled me, that notion. And it worked! And that cracks me up, every time it comes to mind.
Good job, all! When you start your project and need a little dough to nuke a little bad medical debt, hit me up.
We are facing a lot of terrifying problems right now: climate disaster, civil unrest, labor injustice, opioid abuse, the profusion of hate and extremism.
Most problems cannot be solved by throwing money at them. Medical debt is one that can.
I’m collaborating on a project to erase $1 million of medical debt in Michigan.
Fortunately, because of the way this debt is bundled, it’s possible to buy it extremely cheaply on the open market, for just a penny on the dollar (i.e., $1 buys $100 of bundled medical debt). If bill collectors buy that debt, they will hound the debtor for the full value of the debt, even though they only paid 1/100th of that.
But if we buy the debt, we can just forgive it and clear the books for someone who had some bad luck and didn’t have the insurance to cover it.
Incidentally, in Michigan this burden hits Black communities especially hard; buying and erasing debt here is thus a “Good Deed twofer”: one donation relieves a family’s economic woes AND strikes a blow for racial and economic justice.
A bunch of you have already given (we’re over 76% of the way to our goal!), and that’s amaaaaazing! The more you give, the more we can forgive. THANK YOU!
Washington Post: “Play it again, Sam: Inside Bankman-Fried’s last year in the crypto game” by Michael Lewis (gift link 🎁)
Before I listened to this, I had no opinion of SBF, apart from the assumption he was yet another contemptibly oblivious man-child “genius” (and mild embarrassment that he’s a Jew, like me).
When I started listening to this, my contempt was swiftly confirmed. But man, by the end, I was goddamned compelled. Fuck you, Michael Lewis; you win again.
For those interested, in light of Mississippi Book Burning Banning, Seattle Public Library is being fun:
Teens and young adults ages 13 to 26 living anywhere in the U.S. can access our entire collection of e-books and audiobooks. We believe in your right to read what you want, discover yourself and form your own opinions. Fill out the simple form at the bottom of this page to get a Books Unbanned card.
Go forth and read, my darling dears. 📖📚💕
Thunderf00t—who I don’t always agree with or necessarily like—makes an excellent point around the 18 minute mark in this video. The jist of it is this: nuclear weapons are fragile as hell and expensive to maintain (to the tune of ~$1 million per year). If not properly cared for, they don’t kaBOOOOM!🤯 and Chernobyl us all to Mad Max land. Instead, they look perfectly fine and useful, but when you go to launch them, they fizzle like damp firecrackers (albeit damp firecrackers that can spread radiation; but still, while awful, a dirty bomb is not a nuclear bomb).
Russia is a first-order corrupt kleptocracy, as demonstrated by their current Ukrainian misadventure. If they haven’t been maintaining the cheap and easy stuff (like tanks), then they sure as hell aren’t maintaining the expensive stuff that’s hard to spot check for compliance (like nukes).
(Honestly, the whole video makes a series of inter-nesting excellent points, and is worth watching. The major thrust is that there are a lot of dogwhistles blowing out there, signaling who is and is not an actual nuclear superpower.)
This is a “frog boil” story, and may in fact be the perfect frog boil story. If you’re woke-ish, then it is pretty clearly a climate change story. But if you’re on the political right, it may actually seem to much more obviously be an immigration story. It could be a cautionary tale about the dangers of group think (although its up to the reader to determine of its more about anti-mask group think, QAnon group think, CRT group think) or privilege or income disparity.
However you read it, the message is the same: It’s a warning against repeating the same old prayer that humans have repeated prior to disaster for Millenia:
I guess it’s happening, but let it happen in some other neighborhood in some other town far away, above someone else’s roof and out of my sight.
Anywhere but here.
Anywhere else.
Also, absolutely terrific monster-of-the-week. So worth your ears and eyes: PseudoPod 819: “Balloon Season” by by Thomas Ha
Yet another example of how absurdly easy it is to manipulate artificial intelligence—or even just accidentally make them into terrible bigots (and slightly above average antisemites).
From University of Cambridge’s Ross Anderson (via security guru Bruce Schneier, “Manipulating Machine-Learning Systems through the Order of the Training Data”):
Most deep neural networks are trained by stochastic gradient descent. Now “stochastic” is a fancy Greek word for “random”; it means that the training data are fed into the model in random order.
So what happens if the bad guys can cause the order to be not random? You guessed it – all bets are off. Suppose for example a company or a country wanted to have a credit-scoring system that’s secretly sexist, but still be able to pretend that its training was actually fair. Well, they could assemble a set of financial data that was representative of the whole population, but start the model’s training on ten rich men and ten poor women drawn from that set – then let initialisation bias do the rest of the work.
Anderson concludes “It’s time for the machine-learning community to carefully examine their assumptions about randomness.”
I think that’s tangent to the real lesson, which is this:
All machines (including AIs) are created things, and created things bear the biases of their creators in unexpected, but ironclad ways: early color film was shite at photographing people of color simply because the folks who created the color film were all White and unintentionally selected techniques and chemical processes that worked better for their own paler skin tones than they did for darker ones. Similarly, male engineers built crash-test dummies that were roughly their own size and weight—and thus created “safety” features that killed women and children.
